12
FM OPERATOR TYPE-L (OPLII)
APPLICATION MANUAL
CATALOG No.: LSI-2438124
1994.6
a
eed
TABLE OF CONTENTS
1. OPLII
Ps1. COV eL VIEW vce jasseccacesccs vciereas rast gnsivavrnroon ener ssi rteala pean tarot oeteaaneae Saaemenne toate mentee eT
[<2 Features cccscrssssecsssnenesaccsssvcsvsncncanansenstasisseusannonvestentvossasenscnrneasaranerssoatarsonsransen ese ts rer"
1-3 Overview of FM System........ccssescerssssesressssssssnrserncneaecrensanncnenensennesegess asec gern gee"
2. Overview of Functions
9-1] Main Functions .........:.:cesccsssceossscssesnrsssenssocenssenreseansannsoanees sean sesnasorenanencanceanensrenee ees
2-2 Pin Layout |
2-3 Description of Pin Functions
FA Bus Control ccccasssecssssissvososrvaessrsaurssisenssnsanentenenntoncanpnasonannns senbneanaenenarnansnsnsvensnaneres
4.5 Chanmels and Slots ..........:ccccsssscersencenrencenecnseansesssnsssnananaconsaacsenamsenncarensneassngseee i228
9-6 Block Diagram ...........ccssssssceescesecesessesesseeessnscnenssnaseneanaranansnensonesssesesnereeneeeeeeeeeee9
B27 GAPORG TRAD wiscsiesccessuceeorsnovsnanensaiaevsncvinennnnnnnsstino bs tinaniaentpvenacitansenannanarnsn sects ea NFOReest
cryin ce Guislag Giga vale wou utauls sew anapauendaeawedenns sueuagns eoeesbetN EC tesa cae ea site teseerstase meer SS
Sa vidielclus GUR ans sie sin pe¥s Sas cemide cape aesneese emecss eveens seinen esis Seeces see eessen”
Jal Registers ......se-ssesvecsneerrsenecnsssnonvonseveennsensessousnosonenscrecansonsuunrancnnacenenaconennevensensense420e
3-2 Phase Generator (PG) .......--sc-csersnsssssrstnsessescescnevsenssnanenornenesanncssansorvonssannersnneenenens
3-3. Envelope Generator (EG)........ccssectessssreeseseenessnenctsneencensnsnennenssstee snes ee eee eens
3-4 Operator (OP) and Accumulator (ACC) .......cccceeseeeescenesscnsseestenpeneesennncnsannenanans
3-5 Status Information and Interrupt Sigmals..........-sssscereeceerreretetterseesnrereresessr ns
4. Interfacing
4-1 Generation of System ClOCK.........cccssreeseseetersesserstsenarenaennerenseesesenees 2 sass islenmsnailid
4-2 Audio Output Interface
43 Interface to Microprocessor/MicrocOmputer ......-.+sccecssssesreeretsenersercnscersnesseneenersee
saa Wale TEC ede alnie oun Won eed sawsa Bera waseed tap ier nds e ee Sed Sem eee secese sees secewurncersernh try
5, Creation of Music
5-1 Concept of Sound Creation ........ssscserserscsesrenrerersssrneneananenennsnennneserensenen renee re 2 204
5-2 Basics of Sound Creation
ppaev'dunad eda bs cdbegedeusedisucnenuccessnncnegusceenserescovesesseneoueresossonnensnes®
Kuaeu weep eoeaoe Seda We Keene wresheonacsseee Roun sgeeesadeew ime aregeecss ence’ el
5-3 Example of Sound Creation
5.4 Creation of Rhythm Sounds
weeea ony Hae No ouled wan WOE Tea Tes Ts Heme Maw wNeens ees ren kes ees ex eeaconesteseeoseselsn ste
6. Electrical Characteristics ............cscccsesececessseeteeseeseneenessearenssnnsnnennenanansaseesatyg gage
7. Timing Diagrams
Lice heddaadennsGsWseaseaecs aeaceearncanenen ned obauts cundycegesecensisccesedebaseesccsesnanesessec hoses
pe EE EEE TATE
1.
1-1
The FM Operator Type-LII (OPLID) is a new type of sound generator designed for use with Captain
OPLII
Overview
systems and videotext systems. This device uses the same frequency modulation (FM) system used in
our Music Synthesizer Yamaha-DX7 and other similar instruments. This allows for the production
of a wide variety of sounds using software control. This sound generator is also equipped with functions
for the production of rhythm sounds. This does not use the FM system, but instead creates sound
through combining various sound frequencies, including white noise.
The OPLII has also has a built-in low frequency oscillator for vibrato and AM effects, reducing the
amount of programming required to produce special effects.
*
As this sound output from LSI is digital, a D/A converter such as YM3014 is necessary.
Features
FM sound generation system for realistic sound
Mode selection of simultaneous voicing of 9 sounds or 6 melody sounds and 5 rhythm sounds
Built-in vibrato oscillator/amplitude modulation oscillator (AM)
Composite sine wave speech synthesis also possible
input/output TTL compatible
Si-gate CMOS-LSI
5V single power supply
1-3. Overview of FM System
FM refers to Frequency Modulation and is a system using combinations of the higher harmonics
created by modulation. This allows for the generation of waveforms containing high harmonics and
non-harmonic sounds using circuitry which is relatively simple. The correspondence between the
modulation index and spectrum distribution of higher harmonics is extremely natural allowing for
production of a wide range of sounds from natural instruments to electronic sounds.
The following formula (1) expresses the four parameters relevant to FM sound generation.
F = Asin(oct + Isin omt) (1)
Where A is the output amplitude, I is the modulation index, and wm and we are the frequencies of
the carrier and modulator respectively. Formula (1) can also be expressed in the following manner :
F=AfJo(Isin act + J1(1) {sin(oc + om)t — sin(@c — @m)t}
+ J2(I){sin(@c + 2em)t + sin(@c — 2am)t + .....] (2)
Jn (I) is a type 1 Bessel function of the n series. As (2) indicates, the amplitude of each of the harmonics
is expresed by the Bessel function of the modulation index. Formula (1) indicates that FM sound
generation is highly effective for combining special music and sound effects. This is not a string- type
sound source which does not provide an even distribution of higher harmonics. The feedback FM
of this system is shown in (3) below :
F= Asin(oct + BF) (3)
Where 8 .is the feedback factor. This feedback RM is also possible by string-type sound generation
where the higher harmonic spectrum is a sawtooth wave.
The following three function blocks are necessary for FM sound generation :
a. Phase generator (PG) to generate ot
b. Envelope generator (EG) allowing for amplitude A and modulation index to be expressed
as time functions
c. Sin table (sin)
Combining these three elements into a single allows for configuration of the FM system shown in
Fig. 1-1. Thus, if the concept of these units (operator cells : OP) is used, FM sound generation is a
matter of setting the frequency and EG parameters within the units, and then combining the data
between units.
Fi)
b. F(t)=AiSEN ott + A2SIN w2t c. F(t)}=A SIN (at+ BFW)
Fig. 1-1 FM sound generation using unit cells
4. Overview of Functions
2-1 Main Functions
The OPILL is equipped with a total of three voicing modes : nine sound simultaneous voicing
mode, 6 melody/5 rhythm sound voicing mode, and composite sine wave speech synthesis mode. Each
of these modes can be selected by software.
a) 9 sound simultaneous voicing mode :
This mode allows for simultaneous voicing of nine FM sounds having different voices. Both
the rhythm bit (R) and speech synthesis bit (CSM) must he set to mF
b) 6 melody/5 rhythm sound voicing mode :
When the OPLII is set to this mode, the number of melody sounds which be simultaneously
voiced is reduced by three to six, but five rhythm sounds are added (bass drum, snare drum,
tom tom, top cymbal, and high hat cymbal). The bass drum is created using FM sound gen-
eration, the tom tom by sine waves, and the other three rhythm instruments are simulated by
composite frequencies. This mode is effective when connected to Captain or similar systems
using text.
c} Speech synthesis mode :
Speech synthesis using the OPLII is by the composite sine wave speech synthesis method. Voices
are simulated using 3 to 6 sine waves and pitch. :
In addition to the above voicing modes, the OPLH is also equipped with a built-in vibrato oscillator
and amplitude modulation oscillator. These effects can be used to create a sound which closely simulates
the sound of natural instrumnents. Inclusion of these functions allows for a reduction is required
programming. The OPLII also has both a long and short timer allowing for use as reference signals
for scanning of key switches and tempo clock. The short timer can be used as a pitch generator for
composite sine wave speech synthesis.
2-2 Pin Layout
<
2
3
8
z
osy
2
3
4
WR [5
6
7
8
9
—
o
7)
>
D; * NC : Ne Connection
TOP VIEW(24pin DIP, 24pin SOP)
2-3
a)
b)
Description of Pin Functions
oM
Master clock for OPLII
oSY:SH
Clock needed for converting digital output of OPLII into analog signals (PSY) and syncron-
ization signal (SH). These signals allow for direct connection to the YM3014 D/A convetor.
Do~D7
8 bit bidirectional data bus.
CS:RD-WR:A0
Control data bus comprised of Do~ D7.
Write address of register to OPLII
Write contents of register to OPLII
Status of OPLII is read.
Data of data bus Do~ D7 not assured.
Set data bus Do~Dr to high impedance
e)
g)
h)
i)
IRQ
Interrupt signal sent from either of two timers. When the time programmed for the timer elapses,
driven to low level. Interrupts can be masked by program.
IC
Clears the contents of all registers of OPLII, and initializes OPLH.
MO
Outputs music or audio signal converted into 13 bit serial signal. D/A convertor (YM3014
or equivalent) is need for converting this digital output signal into its analog equivalent.
GND
Ground pin
Vee
+5V power supply pin
bd
2-4
Bus Control
Data bus control for reading and writing addresses and data of the registers of the OPLII is
performed by the CS, WR, RD, and AO signals. The following four modes can be set according to
the state of this four signals.
a)
b)
c)
d)
e)
Table 2-1 Mode Selection
Inactive mode
Address write mode
Data write mode
Status write mode
Inhibit
Inactive mode:
When level of CS is “1”, the data bus Do~ D7 has high impedance.
Address write mode :
When an address is to be written, the control signals are set to the address write mode and
the address data is set in the data bus. The address of the designated internal register is set
and data can be written. It should be noted that after address data has been written, 12 cycles
of the master clock (¢M) must elapse before music data is written.
Data write mode :
When the control signals are set to the data write mode, the data of Do~D7 (data bus) is
written to the register having the designated address. A wait period is also required after a
data write, in the same manner as an address write. In this case, the wait period is 84 cycles
of the master clock ($M) before the next data or address is written.
Status read mode :
Setting the control signals to this mode allows for output of the status information stored in
the status register of OPLII. ,
Inhibit :
The data of the bus is meaningless when the control signals are set to this mode. Control of
this data is not possible.
The following precautions must be taken regarding the address and data write modes.
When an address or data is written to an internal register of the OPLII, the following wait period is
required before the next operation is performed. The period varies between the address write and data
write modes. The CPU generates the wait period shown in Table 2-2 for the OPLII. Data integrity
cannot be assured is this wait period is ignored.
Table 2-2 Wait period
ec
Address write mode 12 cycles
Data write mode 84 cycles
Note: The indicated number of cycles for the wait period
is the number of master clock cycles.
- eee ee
oo
2-5
Channels and Slots
The OPLII is capable of voicing 9 FM sounds (9 channels). Each of these sounds has two operator
cells. There is, however, actually one operator cell for the system, and thus the signal must pass through
this operator cell a total of 18 times for 9 FM sounds. The order (slot number) which channel signals
pass this operator cell corresponds to the register numbers. Voicing control of the various sounds ts
possible through control of the registers corresponding to the slots.
The F-number data for each channel! controls two slots.
The relationship between these two slots (first slot and second slot) in the FM modulation mode
is such that the first slot is always the modulating wave, and the second slot is the carrier wave. This
first slot can also be set to the FM feedback mode. Refer to (2-1-9) for settings in this mode.
The relationship between channels and slots is shown in Table 2-3.
2-6
[fo [e[s tela] ss [ro]osfs2] 3/1415} s6] se
a[s[r|zfofets|elelsto|z]s[o]7fel>
eye de dete del
ass [alow coo
ciferlefelcle
Table 2-3 Channels and Slots
Slot number
Channel number
Slot number for each channel
Relationship between data for each slot and
registers (ex. $20~$35)
Relationship between data for each channel
and registers (ex. 3CO~$C8)
es
sees xxx
Block Diagram
AO CS RD WR eM @SY SH TRO
—
6 MO
Do~ D7 Register array
2-7 Address Map
ADDRESS D7 Dé Ds Ds D3 In Di Do
01
Status registers
Dr De Ds Da Di D2 Di_ De
FLG|FLG
Ti] T2
COMMENT
a
TEST DATA OF LSI When the value is “0”, Ds is
DATA GF TIMER-1 compatible with YM3526.
DATA OF TIMER-2
IRQ-RESET/CONTROL OF TIMER-1,2
CSM SPEECH SYNTHESIS MODE/NOTE SELECT
AM-MOD/VIBRATO/EG-TYPE
KEY-SCALE RATE/MULTIPLE
A
KEY-SCALE LEVEL/TOTAL LEVEL
A
ATTACK RATE/DECAY RATE
A
SUSTAIN LEVEL/RELEASE RATE
a
KEY-ON/BLOCK/F-NUMBER
DEPTH(AM/VIB)/RHYTHM(BD, SD, TOM, TC, HH)
FEEDBACK/CONNECTION
ee
WAVE SELECT
on
aa
3. Description of Operation
All functions of the OPLII are controlled by data written from the microprocessor to the register
array. The shape of the envelope for music, modulation factor, frequency, voicing, mode, and other
parameters are determined according to the data which is written to the registers. Data can be combined
to generate the sound of a piano, violin, or other instrument. There is an extremely large number of
combinations with a high degree of complexity. This chapter deals solely with the function of the various
registers. Refer to the chapter on creating music for details on the various possible combinations.
3-1 = Registers
The registers comprise an area of 777 bits as shown in the address map of 2-7. The addresses in
this diagram are the subaddresses allocated to the various registers in the OPLII. Music data is written
to the internal registers through these subaddresses. Thus, data is stored in the OPLII by first sending
the subaddress data which will hold this data, and then sending the music data itself. When the same
subaddress is to be accessed a number of times, the subaddress data need only be sent in the beginning.
Music data can then be sent, without address data, to update the data.
The initial setting for all registers is “Q” (initial clear = “O”). )
3-1-1 Test : Address ($01)
The only use of this address is for testing of the LSI device by Yamaha.
The bits are normally all “0”. The D5 bit, however, has a special meaning. The output waveform can
be controlled by setting this bit to 1. (Refer to 3-1-2)
3-1-2 Timer
There are two timers: Timer | which has a resolution of 80ps and Timer 2 which: has a resolution
of 320s. Starting, stopping, and flag contro! of both timers is possible. When a timer flag is set, the
IRQ pin is driven to low level, and the microprocessor is notified of a timer interrupt.
i) Timer-1 : Address ($02)
Timer 1 is an 8 bit presettable counter. If an overflow occurs, the flag 1s set, and the preset 3
value is loaded. In addition to normal timer functions, Timer 1 is also used for control of
composite speech synthesis. When an overflow occurs in this mode, all slots are set to Key-ON
(voicing), and then immediately to Key-OFF. This operation allows for composite speech
synthesis.
$02 D7 De Ds Da D3 D2 Th Do
Tov(ms) = (256-N1)+#0.08 @eM =3.6MHz
Ni= D702? + De*26 +...... +Dit2+Do
ii) Timer-2 : Address ($03)
- ‘Timer 2 is an 8 bit presettable counter like Timer 1. The diference between the two timers is
that the resolution of Timer 1 is 80ps, and the resolution of Timer 2 is 320ps.
$03 D7 De Ds Da Ds D2 Mi Do
Tov(ws) = (256-N2)#0.32 @oM = 3.6MHz
No = D7027 + Dor2* + |... +Di+#2 + Do
iii) Timer Control : Address ($04}
This register is used for start, stop, and flag control of Timers ! and 2. These operations are
controlled by the bit state of Do, Di, Ds, De, and D7.
&
Me
ws
<
2
Do (STi) : Controls starting and stopping of Timer 1.
When this bit is “1”, the preset value is loaded into Timer 1, and counting started.
When this bit is “0”, Timer 1 does not operate.
) D1 (ST2) : Performs the same operation as Do (ST1) for Timer 2.
Ds (MASK T2) : When this bit is “1”, the flag for Timer 2 is masked (always “0”), and has no
effect on operation of Timer 2.
De (MASK 11) : This bit masks the flag for Timer 1.
) D7 (IRQ RESET) : Resets the flags for Timers | and 2.
When the D7 bit is set to “1”, the data of Do~Ds is ignored, and the D7 bit is auto-
matically reset to “0”.
3-1-3 CSM Mode/Keyboard Split : Address ($08)
This register sets the mode to the music mode or speech synthesis mode, and determines the keyboard
split for keyboard scaling of rate.
$08 De|Ds D4 Ds D2 Di Do
NOTE SEL
De (NOTE SEL) : This bit controls the split point of the keyboard. When “0”, the keyboard split
is the second bit from the most significant bit (MSB) of the F-Number. The MSB of
the F- Number is controlled when “1”. This is illustrated below.
Octave
Block data
F-Num’MSB
F-Num‘2nd
Keyboard split number
Octave
Block Data
F-Num‘MSB
F-Num2nd
Keyboard split number
* DON’T CARE
D7 (CSM) : The composite sine wave speech synthesis mode is selected when “1”. All channels
must be in the Key-OFF state when this mode is selected.
3-1-4 AM/VIB/EG-TYP/KSR/Multiple : Address ($20 ~ 35)
This register controls the multiple for the conversion of the frequency data given by the envelope
shape and F-Number into carrier and modulating wave frequencies corresponding to the frequency
components of music.
$20~3$35
fos [b:[b [dsb D:_De
11
Do~Ds3 (MULTIPLE) : The frequencies of the carrier and modulating waves are controlled by the
multiples shown in Table 3-1.
} < Example >
Frequency of F-Number of
Multiple for carrier wave
Multiple for modulating wave 7
F(t)=Esin (oft + Isin(7oft))
—
Table 3-1 Multiples
MUL | Multiple | MUL
4
Multiple
Multiple
oo
Ds (KSR) : This bit gives the key scaling for the rate.
The leading and trailing edges (attack and decay) of the sound of natural instruments
tends to increase as the interval becomes higher. Key scaling of the rate allows for a
simulation of this phenomenon. Table 3-2 shows the value of the offset which is added
to the speed for each the intervals. Thus, the actual rate for ADSR (attack, decay, sustain,
) and release) becomes the rate set for each with the indicated offset added.
RATE= 4#R + Rks.
© R is the value set for ADSR
© Rks is the key scaling offset value
© The RATE=0 when R=0
* Nis the key scaling number
12
Ds (EG-TYP) : Switches between connecting sound and diminishing sound.
A diminishing sound is selected when Ds=“Q”", and a connecting sound is selected
when “1”. The difference between these two voicing modes lies in the use of the release
rate. This is illustrated in Fig. 3-1.
Ds =“0” Diminishing sound Ds=“1” Continuing sound
DR DR
SL SL
Key-ON Key-ON Key-OFF
AR=ATTACK RATE DR=DECAY RATE
$L=SUSTAIN LEVEL RR=RELEASE RATE
Fig. 3-1 Two Types of Voicing Modes
De (VIB) ; ON/OFF switch for vibrato. Settting this bit to “1” causes a vibrato effect to be applied
to the slot. The frequency for this is 6.4Hz (@ ¢M=3.6MHz), and the depth of the
vibrato is determined by VIB-DEPTH of the BD register.
D7 (AM) : ON/OFF switch for AM modulation. Setting this bit to “1” causes AM modulation to
be applied to the slot. The frequency for this is 3.7Hz (@ oM = 3.6MHz), and the depth
of the vibrato is determined by AM-DEPTH of the BD register.
3-1-5 KSL/Total Level : Address ($40 ~ $55)
The total level is used to control the modulation factor (voice) and volume for the output of the
envelope generator. Level key scaling (KSL) is similar to rate key scaling. This allows for simulation
of the tendency for the output level of natural instruments to decrease as the interval increases.
[D7 Ds|Ds D4 Ds D2 Di De
Do~Ds (Total Level) : The minimum resolution for attenuation is 0.75dB, and the volume can
attenuated by a maximum of 47.25dB.
$40~$55
13
Tabte 3-3 Total Level
ee
Ds Ds D3 Dz Di Do
24 «12 6
F Degree of attenuation 3001S 0.75
dB dB @B dB dB 4B
De~ D7 (KSL) : Bits for control of level key scaling.
The key scale mode attenuates the level as the interval increases. The degree of atten-
uation can be set as 0 (no attenuation), 1.5dB/octave, 3dB/octave, or 6dB/octave.
Table 3-4
Degree of Attenuation
RE ne fr eee sae naa ee eee ------- I------- =I-----+-- 1
mr pr bt or 2 8 3 + 4 « S&S t 6 2 F 3
} jo a oe a oe oe ee wf @ 2.8% 4
[on--NM----- ey o---5-+- teen taelanaaae SS Ceca |
I 0 1 9.008 T 0,00 I oud T o.00% To. 000 7 9.000 I o.000 1 9,000 T
I 0,000 IT 0,000 1 0. ean IO, Oboe 1 o.oco T Of. 000 I ooo) L 6,000 7
es Ce ais Oe ee er. I aE AE
1 | I 0,000 1 9,000 T 0.000 IT o,000 T G00 T 5.000 1 0.000 1 6.900 I
I I 0.9000 1 0.750 I |. 125 I 1-800 1 1.875 1 2,250 § 2.6275 I 3.000 I
[onna- lan enn ns Coraline Cocaine [----46--[e--- a |
t 2 I o,.000 1 «9.000 T o.000 1 o.006 T 0,606 I 4.125 1 1.875 1 2. 625 I
I I 3.000 I 3.750 I 4. 125 1 4.500 1 46.875 1 5.250 1 S.625 1 6.000 I
I----" I+------~" l-----" wejyerrcee [oss foes 35 t- | -20 een lane I
I I o.ooo LL G.oun 1 000 L 1.975 1 3.000 1 & 125 1 4.875 1 5S. 625 1
1 3 1 §.000 1 6.750% 7.125 1 7.500 1 7.875 1 6.250 1 @,625 1 3.000 I
I----- 1-------- {----+--- [-------- ta----4n- I-------" 1-------- Jo-w =n [-------~ 1
) I I vw. ogg To, bod xy 3.000 JT 4. B75 1 6.008 . 7.225 1] 7.075 1 8.625 [
1 4 2 3.000 1 9.750 1 10,125 71 10. 500 110,875 1 11.259 1 11,625 1 12,000 1
1----- J-<------ Y+------- [-----~-- [-------- I-------~- [~------- I-------- I-------- I
I 5 I Ooo TL B.mo 6.000 © 7.875 TF 3,000 1 10,1285 § 10.875 1 11.625 1
1 1 12.000 I 12.750 T 15.125 113.503 1 17,975 1 14 250 1 14.625 I 49.000 I
\----- [-~<----—- iow [-------- \ernenco- l--+----~ leer foro [oe--- == t
1 5 I 06.000 I Bond Y B.ueo To id. @75 1 12.000 1 13.125 1 1Z.975 1 14.625 I
I t 15.000 1 15. 750 416.125 1 16.500 $ 16.875 1 17.258 | 17.625 1 18.a00 I
F aecombaataal lernrerr t--<2---~-" ]-------"- {--"<e<c= [oseeres— {-------- \------7- |----c--r" I
I 7 1 o.000 F 3. ooo To ized lo ls. p75 1 15.000 116,125 1 16. 975 1 17.625 !
I 1 18.00 1 18.750 i 49. 125 1 19.500 1 19,875 7 20,750 1 20.625 1 21.000 I
1----- t-------" J--"----- {---e--77 , a aaaian a Lesa meiner Lees ated {-corcce i
Unit : dB
Note: + F-Number is the value of most significant four bits
» 1.54B is 1/2 of above
» 6dB is twice above
14
3-1-6 Attack/Decay Rate : Address ($60 ~$75)
The attack rate sets the rising time for the sound. The decay rate is the diminishing time after the
attack. The time settings for each of these rates are shown in Table 3-6.
$60 ~$75
D7 De Ds D4a| D3 D2 Dm hm
AR DR
oP 2F) Ze ae ae” ea ae
3-1-7 Sustain Level/Release Rate : Address ($80 ~ $95)
For continuing sounds, the sustain level gives the point of change where the attenuated sound in
the decay mode changes to a sound having a constant level. For diminishing sounds, the sustain level
gives the point where the decay mode changes to the release mode.
For continuing sounds, the release rate defines the rate at which the sound disappears after
Key-OFF. For diminishing sounds, the sustain level indicates the attenuation prior to reaching the
sustain level (point of change) while the release rate indicates the attenuation after the sustain level
is reached.
$80~S$95
D7 De Ds D4} D3 D2 Di Do
DR
dB dB dB 4B|2* 27 2' 2°
“93dB when bits Da~D7(SL) are all “1”.
* The attenuation time for the release rate is the same as
that shown in table for decay rate.
15
Table 3-6 Attack and Decay Times for Various Rates
The rates indicated below are those after key scaling. The rate value is obtained by taking
the most significant four bits (RM) and subtracting the least significant two bits (RL)
(RM-RL). RATE = RM*4 + RL
eee EG ATTACK TIME +4 wee EG DECAY TIME *0% eee EG ATTACK TIME #44 «ee EG DECAY TIME »**
RATE ms RATE ms RATE ws RATE pee
(@ 10% - 90% } (@ OdB - 36GB }
is = is 3 0. 00 is 3 2.40
15 3 0.00 15 3 o.51 :
1% z 0. 00 15 2 2,40
15 2 D.00 15 2 0.51
, 151 0.00 151 2.40
{51 0, 00 iS 1 o. St .
race aes ans aoa 15 6 0,00 15. 6 2.40
14 3 O11 14 3 0.58 aie o. 20 14 5 2.74
1a 2 O.13 14 2 0.63 14 2 +26 14 2 3.20
ie ees . ee oe 14.1 o.3e 14 1 4
. 140 Oo. 28 14 0 4.80
44 0 0.19 14 6 1.01 ; :
4 13 3 O42 13 3 5.48
13 3 0.22 1% 3 1.15 .
13 2 0. 26 1s 2 6.40
13 2 O. 26 13 2 1.25 :
: tz 1 a, S6 131 7.68
131 o.731 131 1.62 a
> 13.4 0. 7G 13 0 9.66
120 0.37 13.0 2.02
iz 3 0.80 12 3 10. 96
123 0. a3 12 3 2.32
12 2 0. 92 12 2 12. 80
iz2 0.43 iz 2 2.68 :
tz 1. tz 1z 1 15.35
121 0.61 iz 1 3.22 5
iz a 1.40 12 0 13. 20
iz 0.73 iz 0 4.02
‘ ik s 1.56 11 3 21.92
11 3 0. 8S 11 3 4.62 Z
11 2 1.84 11 2 25. 56
ii 2 0. 37 41 2 5.28
111 2.20 42 4 30.68
14 1 1.45 ait 6.42
110 2.76 11 0 38. 36
11 0 1.45 110 a,c
10 3 3.12 10 3 43. BA
10 3 1.70 10 3 3.24
10 2 3.6B 10 2 51.12
id 2 1. 9% 10 2 10. 76 ; 5
1% 2 4.40 1G 1 61.36
10 1 2.26 10 1 12.64
19.9 S. 52 10 0 76.72
40 0 2.99 19 0 15.04
93 &. 24 a3 87.68
33 3.39 33 15.48
; 92 7.76 32 102. 24
92 3.87 22 21.52
91 a. 90 $1 122. 72
31 4.31 31 25.68
30 11.04 30 153.44
30 5.79 a0 22.08 2
° a3 12. 4B BS 175.36
et 5.78 BS 56.36 = a
a2? 14, 72 B2 204. 48
a2 7.74 a2 43, 04
Bi 17.60 Bi 245.44
a1 3. 012 al 31.3
80 22.08 Bo 306. 6
80 11.58 50 Ba. 16 -
7% 13.57 mae 73.42 cae 24. 36 73 350.72
: z 72 23. At 72 408, 96
72 15.49 72 BE. 08
: 71 35. 20 71 490. 89
74 19.05 74 102.72 if
oh ae 8 eae 70 46.16 70 G13. 7€
23. abe & 3 49. 92 & 3 701.44
63 27,14 &3 147.84 si
& 2 59, 688 6 2 @17.92
& 2 30. 98 62 172.16
6 i 36.10 6 ti 205.64 pode Fn Se 61 981.76
wi 205. Bo BB. 32 6 0 1227.52
& 0 46.54 ran) 256. BG
ides 53 99. B4 53 1402.88
53 $4.27 5 3 295.68
52 51.35 sz Baa. 32 ae eis S2 1655. Be
51 72.13 51 410. BB = 4 140. BO 3 1963.52
50 176. 64 50 2455. 04
so 92.67 50 513.268
is inaicBe ae en xe 43 199, 68 43 2805. 76
Rs igen ne oan Bi 42 235. 52 42 3271.68
a1 144.38 a's 621.76 a4 281.60 41 3927.04
a 4 40 352.28 40 4919. 068
a0 185.34 40 1026. 56 ae cae tk
33 217-03 a: 1182.72 : ve ad 5611.52
ee ge = 5 jaan oh 32 471.06 32 5543. 36
31 288.77 34 1643.52 age S63. 20 31 7654. 08
30 370.63 30 2053.12 ie 705-56 “f eS
——_ ces th =e aga 2 798, 72 23 11223.04
22 495.62 22 2754.56 a $42.08 22 15006. 72
21 577.54 21 3287.04 =. 1126.40 21 {5708.16
zo 74ai.38 20 4106, 24 20 1413.12 206 19640, 32
is 268.35 t > 720,98 is 1537.44 13 224466. 08
12 991.25 i< S504. 12 1¢ 1684.16 12 26173.44
ti 115%, 07 41 6574.00 11 2252. BO a. 31416.32
10 1482.75 10 B212. 48 io 2826. 24 10 =~%39260.64
Note: There is no change in the envelope when the rate is “0”.
3-1-8 Block/F-Number : Address ($A0 ~ $B8)
Data for determining the interval and scale. The F-Number is determined by both the $A* and
$B* registers.
$A0~$A8 ©|D7 De Ds Da D3 Tr
F-Number
QP 9
$B0 ~$B8
Do~D7($A*), Do ~ D1 ($B*) (F-Number) : The F-Number is indicated by the total of ten bits formed
from the 8 bits of the $A* register and lower two bits of the $B* register. The data
of the F-Number gives the scale. This value is determined by the procedure outlined
below.
D2~Da4 (BLOCK) : Gives information on the octave.
Ds (Key-ON) : Bit corresponding to ON/OFF of the keyboard. When this bit is 1”, the channel
is ON and voiced. Key-OFF when “0”.
* F-Number/Block
With the OPLII, the required frequency can be obtained by giving the phase increment corre-
sponding to this frequency. This increment is determined by the F-Number, Block, and Multiple in-
formation.
First, the increment of the desired frequency is obtained. This is given by the formula shown below.
AP =fmus*2!9/fsam fsam = fM/72 a TD
fmus_ =: desired frequency
fsam : sampling frequency (SOkHz)
{M : frequency of input clock (3.6MHz)
The above allows for the phase increment to be obtained. As many of bits would be required to
express this value, only the data for a single octave is given, and this increment is shifted for each
octave (2x, 4x .....). This aliows for expression of the increment in the following manner.
17
More
AP = 284F’*MUL
B : octave information
r : increment limited to a single octave
MUL : Multiple data
Formulas @ and @ permit the increment (F’) to be expressed in 10 bits. The F-Number and
Block can be expressed as follows.
F =(fmus*2!9/fsam)/2°-*
F : F-Number data
b : Block data
Table 3-7-1 F-Number (1)
Frequen
sconce F-Number
{4oct)
; 363
LThZ
1
1 1
=e ew et OG Seo & &
\
I
I
i
I
1
\
|
'
|
I
t
I
!
\
|
\
|
!
— = © 2 CoO OF mere =
Frequency
(4~ Soct)
@MUL=1
oor, wf OS SC
[el
a
ws oOo So fo +
eerooerwoor ee 9
0
0
1
a
1
0
1
—_— & ee &
ceconmpeocooaesecemlmr nt
aecocoreroes ec 9 ©
$A*
Da Ds
0
1
0
I
0
U
0
0
0
0
1
—
eooeocoomdtroo ¢
mw = OS ocr oso
ewe ep ODS SF
oeroooo -
-—_e wre ooeoewerseosdes-
De Di
- ooo = Fe ee me OO
$B* ! $A¥
Di Dol D7 De Ds Ds Da D2
6 1 1 0 1 0
Di Do
]
Do
eoocor ss?
if
id:
| }
3-1-9 Feedback/Comnection : Address ($CO~$C8)
This register determines the modulation factor for self-feedback and the type of FM modulation.
CONNECTIO
Do (CONNECTION) : Connection controls the manner in which two slots are connected. The FM
modulation mode is selected when the data is “0”. The two slots are connected in parallel
for sine wave output in the composite mode when the data is “1”. i
yee Fe
—
OUT
Di~Ds3 (FEEDBACK) : Gives the modulation factor for feedback FM modulation of the first slot.
Table 3-8 Modulation Factor
Cece Ft ae ee
19
3-1-10 AM VIB-Depth/Rhythm : Address ($BD)
This allows for control of AM and vibrato (VIB) depth, selection of rhythm mode, and ON/OFF
control of various rhythm instruments.
ole ag op lim a i
Do~Ds (RHYTHM) : The OPLII is set to the rhythm sound mode when Ds=“1”. Channels
7 ~9 (refer to page 9} are the channels for rhythm sounds. Thus, music (melody section)
! is limited to 6 sounds. Do~Ds allow for ON/OFF control of the various rhythm in-
struments. This means that the $B6, $B7, and $B8 Key-ON registers must always be
“9”, Slots [3~18 correspond to the rhythm sounds shown in Table 3-9. Data such as
rate must-be input as a value appropriate to each rhythm sound.
Table 3-9 Rhythm Slot
De (VIB-DEPTH) : The vibrato depth is 14 cent when De =“1", and 7 cent when “O's
D7 (AM-DEPTR) : The depth for AM is 4.8dB when D7=“1”, and 14B when D7= “0”.
3-1-11 Wave Select
When bit Ds of address $61 is “9” the OPLII is fully compatible with YM3526 (OPL); there are
no differences between the two devices. If a sine wave is input in this mode, the output will be a sine
wave like the input. When bit Ds of address $61 is “1”, the input sine wave will be output as the distorted
wave shown in Table 3-10.
D7 De Ds Da Ds D2
ea
$E0~SF5
Table 3-10 Wave Select
ages
3-2 Phase Generator (PG)
The phase generator is a circuit which obtains a phase value corresponding to the required frequency
through accumulation of the increments for each unit of time. This increment is created from the
frequency information (F-Number, Block, and Multiple) sent from the registers. This circuit is also
equipped with a vibrato generator allowing for creation of a vibrato effect through combining the
output of this generator and frequency information.
3-3 Envelope Generator (EG)
The envelope generator controls the rates for attack, decay, and release, sustain level, total level,
etc. These parameters give the changes in voice and level which occur over time. The dynamic range
of the envelope generator is 96dB (resolution of 0.1875). Indication for the envelope generator is
logarithmic or in terms of degree of attenuation. The basic envelope shape is shown in Fig. 3-2, The
special characteristics of this shape is that the change in level during the attack is exponential while
it is linear during the other sections of the envelope. The crossover from attack to decay occurs at
the OdB point, and the decay changes to sustain when the sustain level is reached. Release begins when
at the Key-OFF point. The effects of total level, level key scaling, and amplitude modulation are added
to the envelope generator for changeing the shape of the envelope. ,
Fig. 3-2 Envelope Waveform
21
3-4 Operator (OP) and Accumulator (ACC)
The operator is a circuit for FM calculation. The operator calculates the SIN value based upon
the phase output from the phase generator, and this is combined with the output of the envelope
generator. If the result is the modulating wave, it is sent back to the input of the operator. If music,
the output is sent to the accumulator. The data for feedback and connection controls this transfer.
The accumulator accumulates the operator output for the various channels. The results of this
calculation are converted to offset binary data consisting of a 10 bit mantissa section (including sign)
and 3 bit exponent section. This data is output from the LSB as shown in Fig. 3-3.
Dt D2 Ds Ian Ds ps D7 Ds
‘SIGN
Fig. 3-3 Output Timing
Internal data of OPLH MO output data
Sign 15 14 13:12 1 10 9 8? ¢ i aes 2
w
wa
ifs)
Ps)
“A
=
é
“>
f--)
~1
a
w
a
w
i
-
:
a
ie ki e ee ee
)
l
|
icc. Ss 627 eee), EIIcSeeeee es!
Fig. 3-4 Internal Data and Output Data
N
LN
3-5 Status Information and Interrupt Signals
The two timers of OPLII are capable of setting flags according to a set period. These flags can
be read as status signals or interrupt signals. Thus, the load on the CPU can be reduced by using the
timers as tempo counters or for generating keyboard scanning signals. The interrupt signals have an
open drain allowing for linking to other chips.
Status signal
Ds (Timer 2 FLAG) : Flag signal set by Timer 2. Set to “1” when Timer 2 reaches the set time.
This data remains until reset.
De (Timer 1 FLAG) : Same operation as Ds for Timer 1.
D7 (IRQ) : Set to “1” when either Ds or De is “1”.
23
CPU
CPU
Preamp
a 4.7pF
1.8KQ
28D655
Fig. 4-2 Audio Interface
43 Interface to Microprocessor/Microcomputer
Do~D7 of OPLH is a bidirectional bus allowing for an interface to a microprocessor. Address
data and status information is passed between the two devices over this bus. CS, RD. WR, and AO
are bus control signals for managing the transferred data. This allows for creation of an FM sound
generation system using the smallest configuration of the OPLII, memory, and a microprocessor.
Data bus
Control
1/0 control
signals
Fig. 4-3 Interface to CPU
25
2 AANA li
5, Creation of Music
This chapter deals with the data which can be input to the registers of the OPLII for creating piano
or brass instrument music.
5-4 Concept of Sound Creation
The basic concept behind FM sound generation is to first fully understand the characteristics of
the desired instrument. For example, the envelope of piano sounds is such that there is a sharp attack
when the keys are pressed, and then the sound gradually disappears as the key is held down. There
are also a large number of harmonic overtones during the attack with this number decreasing over
time until a nearly constant harmonic configuration is attained. After such characteristics are un-
derstood, the means of attaining this sound through FM sound generation can be considered. The
output amplitude can be determined from the envelope characteristics with the harmonic structure
determining the modulation exponent. As this structure of harmonic overtones is related to the fre-
quency of the operator, the frequency ratios can also be determined to a certain degree. In this manner,
the characteristics of the desired sound roughly determine the FM parameters. Then, the various
parameters are adjusted while listening to the sound, until the desired voice is obtained.
5-2 Basics of Sound Creation
FM sound generation uses effects obtained by using a modulator to modulate the carrier. Thus,
the pitch, tone, and level of the music can be determined by skillful manipulation of the basic FM
parameters (carrier output level, modulator output level, feedback level of modulator, frequency of
cartier, and frequency of modulator). The relationship between these parameters and the parameters
of the OPLII is shown in Table 5-1.
i) FM connection (CONNECTION = “0”)
All of the FM characteristics shown in Table 5-1 can be expressed. As operator 1 is equipped
for self-feedback, combination with operator 2 allows for a two stage FM connection for high
harmonic output.
ii) Parallel connection (CONNECTION = ay
The sum of the two operators is obtained with operator 2 always generating a sine wave. Thus,
harmonically shifting the frequency of the two operators allows for effects such as the coupler
effect of a pipe organ. In addition, operator 1 is equipped with the same feedback capabilities
as described above for the output of higher harmonics.
5-3 Example of Sound Creation
Table 5-1 Basics of Sound Creation
[item Applicable parameters MIN —(change in sound)~MAX
Output level of carrier TOTAL LEVEL Min, level -————~ Max. level
Data for A, D, $, and ae f
Output level of modulator Key Scale data Round sound ~——- Bright sound
harp fone
Feedback level of modulator a. ee Normal tone~———~>_ (Noise)
Frequency of carrier MULTI Low pitch--_—— + High pitch
| Frequency of modulator (BLOCK/F-Number) Near harmonics ~-Far harmonics
26
ee |
i)
(i)
(2)
(3)
(4)
(5)
(6)
ii)
()
(2)
(3)
(4)
(5)
Electric Piano
Selection of connection
The connection is set to “0”. Almost all voices can be obtained from this connection. This
allows for both accenting on the attack of operator 1 and for rich high harmonics.
Determining the frequency for operators .
The MULTIPLE for both operators are set to “1” in order to obtain higher harmonic frequencies
which are integer multiples.
Output level of operators
The output of the modulator is altered to adjust the tone- When determining the level for
operator 1, the bass is first set so that the sound is like a piano with a rich range of higher
harmonic frequencies, and then the change in the treble is adjusted by level scaling for operator
1. Level scaling for the treble is required until the output is almost a sine wave.
EG setting
The next step is to determine the envelope for level and timbre. First, operator is set for an
envelope which provides a sharp attack but a relatively prolonged sound. Operator 1 which
forms the modulator is set so that there are a great number of harmonic overtones during the
attack followed by a constant timbre with no changes. Key scaling is also used for operator
2 to provide level adjustment. Rate scaling of the treble should be used for a sharp sound.
Readjustment of data
The procedure for sound creation is nearly complete. The timbre can be altered to a certain
degree by the envelope generator settings, etc. Here, the output level of the operator and the
feedback level are readjusted to touch up the sound. For example, if you feel that the metallic
echo is too strong, the level of operator 1 can be reduced.
Adding effects
Finally, a tremolo effect is added using the LFO to create a sound which is like an electric
piano. This can be done using the amplitude modulation (AM) function of the OPLI, or by
programming the total level to be updated at a period of 2 ~6Hz (use of a triangular wave is
possible).
Trumpet
Selection of connection
The connection for brass instruments is also “0”. The bold brass sound of a trumpet can be
created by adjusting the feedback level of operator 1.
Operator output
The total level for operator 1 (modulator) is set to 4 moderate value in the range of $10 to
$28. The feedback level should be set to the maximum level of 7 for bright reverberation.
Frequency of operators
Basically, both operators should be set to a multiple of 1.
EG
There should a slow attack for both operators. For brass sounds, the attack of the modulator
comes completely after the attack of the carrier. This allows for the characteristics attack of
brass instruments to be captured.
Key scaling
Clarity will be lacking in the treble due to the slow attack of the envelope. Slight rate scaling
is needed to prevent an unnatural sound when playing fast passages.
27
(6) LFO
No matter how well a brass instrument is played, the pitch will quiver during long tones. A
vibrato effect should be added to express this.
5-4 Creation of Rhythm Sounds
Rhythm sounds are created using channels 7, 8, and 9. These three channels (6 slots) can be used
to create a total of five sounds. The bass drum (BD), however, uses two slots for generation of FM
sounds. Thus, a bass drum sound can be created using the same basic procedure as described in
(a)~(c). This explanation will deal only with the remaining four sounds: high hat, top cymbal, tom
tom, and snare drum.
The OPLII is equipped with a noise generator which allows for combining a number of frequencies
and a white noise generator for creating rhythm instruments. This noise generator uses 8 and 9 channel
frequency information (Block/F-Number/Multi), the proper phase output for various rhythm in-
struments is output by combining this signal with white noise. The resultant signal is sent to the
operators. Experience has shown that the best ratio for the two set frequencies is 3: 1 (f?7ch =
3*f8ch). So far, the phase data for the various instruments has been obtained. Finally, envelope in-
formation is combined with this output. As the envelope is set for one slot which corresponds to a
single rhythm instrument, values which express the characteristics of the instrument are set in the
parameter registers, in the same manner as melody instruments. (Refer to 3-1-10).
The above allows for the generation of various rhythm sounds.
28
6. Electrical Characteristics
1. Absolute Maximum Ratings
Pin voltage
Operating ambient temperature
Storage temperature
2. Recommended Operating Conditions
Fripiet aio
5.5
0
Address hold time
Chip select write width cS
Chip select read width cs
z
Write data setup time
Write data hold time
Read pulse width
Read data access time
Read data hold time
Output rise time
Output fal! time
Reset puise width
Kfe
i Fig. A-I Clock Timing
Fig, A-2. Write Timing
Do~ D7
Fig. A-3 Read Timing
30
4, Timing Diagrams (Timing is based upon settings of Vin =2.0V and ViL=9.8V)
Note: fcsw, tww and twou are based on either
C5 or WR being driven to high level.
Note: tacc is based on whichever of CS or RD
goes to the low level jast.
tcsr, trw and trpu are based on either CS
or RD being driven to high level.
oM
tor tor tor2
Fig. A-4 6M and oSY Fig. A-5 0M and SH-MO
Fig. A-6 Reset Pulse
31
g. Package Dimensions
(1) YM3812
Notch or i-pin index mark
(2) YM3812-F
2,00+0.15
B.40+0.20
11.B80+0.40
x
<
=
°
oe
a
0.10+6.10
DIMENSIONS IN MM
32
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11